open research onlineoro.open.ac.uk/172/1/spar_holwell.pdf · soft systems methodology: other voices...

Open Research OnlineThe Open University’s repository of research publicationsand other research outputs

Soft systems methodology: other voicesJournal ItemHow to cite:

Holwell, Sue (2000). Soft systems methodology: other voices. Systemic Practice and Action Research, 13(6)pp. 773–797.

For guidance on citations see FAQs.

c© [not recorded]

Version: [not recorded]

Link(s) to article on publisher’s website:http://dx.doi.org/doi:10.1023/A:1026479529130

Copyright and Moral Rights for the articles on this site are retained by the individual authors and/or other copyrightowners. For more information on Open Research Online’s data policy on reuse of materials please consult the policiespage.

oro.open.ac.uk

http://oro.open.ac.uk/help/helpfaq.html

http://oro.open.ac.uk/help/helpfaq.html#Unrecorded_information_on_coversheet

http://oro.open.ac.uk/help/helpfaq.html#Unrecorded_information_on_coversheet

http://dx.doi.org/doi:10.1023/A:1026479529130

http://oro.open.ac.uk/policies.html

Systemic Practice and Action Research, Vol. 13, No. 6, 2000

1094-429X/ 00/ 1200-0773$18.00/ 0 2000 Plenum Publishing Corporation

773

Soft Systems Methodology: Other Voices

Sue Holwell1

Received July 30, 2000

This issue of Systemic Practice and Action Research, celebrating the work of PeterCheckland, in the particular nature and development of soft systems methodology(SSM), would not have happened unless the work was seen by others as beingimportant. No significant contribution to thinking happens without a secondaryliterature developing. Not surprisingly, many commentaries have accompanied theongoing development of SSM. Some of these are insightful, some full of errors,and some include both insight and absurdity. Checkland (1999, p. A42) opines, inthe recently published 30-year retrospective, that “SSM has been ill-served by itscommentators.” Scrutiny of the secondary literature on SSM provides support for thisview and also identifies some general characteristics and trends that are important tothe development of SSM and, incidentally, reinforces some existing conclusions.

KEY WORDS: soft systems methodology; Peter Checkland.

1. INTRODUCTION

Selecting a base bibliography for a review of these “other voices” is far frombeing straightforward. References to SSM and citations of Checkland’s work arefound in the literature of many disciplines, as diverse as: agriculture, geography,and nursing, for example, as well as management and IS. Here, pragmaticallyselection has been limited to the management (including systems) literature ongeneral problem-solving and the IS literature.

Given this limitation, you would expect SSM to be treated unambiguously,particularly as many are would-be straightforward expositions of the work. Thisis not the case. The management and systems literatures include references toboth SSM and Checkland’s work on SSM in IS, and the IS literature includescontributions that only describe SSM as a problem-solving approach (even in IStextbooks).

The 250-item References section used here is drawn from a wide range of

1Cranfield University, RMCS, Shrivenham, UK.

Holwell774

journal papers, conference papers, and textbooks. In it there are of four kinds ofreferences:

(1) general citations (many of which are “honorific” in the sense that theyare intended to recognize the whole body of work);

(2) student texts [inclusion of SSM in student (particularly IS) texts is arelatively recent phenomenon (e.g., Harry, 1994; Mason and Willcocks,1994; Pidd, 1996)];

(3) specialized journal articles (papers more specifically concerned withSSM); and

(4) specialized IS books [a small group of books, strongly influenced andinformed by Checkland’s work (e.g., Avison and Wood-Harper, 1990;Davies and Ledington, 1991; Jayaratna, 1994; Lewis, 1994; Stowell,1995b; Stowell and West, 1994)].

The References are considered as a whole and the critical exegesis that iden-tified the arguments, concepts, and language which have been noticed outsideLancaster and how they are understood is described in detail elsewhere (Holwell,1997). Where references are cited here, they are by way of example or illustra-tion and may not necessarily reflect the overall view of those authors. They arenot intended as being definitive of either type or content.

Some general observations are relevant. First, there are two very differ-ent levels of discussion. One is a basic level that broadly covers Checkland’swork as it is described in Systems Thinking, Systems Practice and Soft SystemsMethodology in Action. These accounts are found in textbooks or are explana-tory overviews in journals that do not have a history of discourse on SSM, suchas Cavaleri’s (1994) account. One obvious aspect is that most accounts not onlyuse the same language and concepts as Checkland but also closely follow thelogical structure of his account. Such close paraphrasing of Checkland’s workmeans that they are broadly only repetitions of it. But this is not surprising giventhe purpose that such expositions serve.

The other discussion is at a more sophisticated level. Here much of Check-land’s work is taken-as-given and the focus is on particular aspects [for example,the multiple uses of Weltanschauung (Fairtlough, 1982), modes of SSM use inAustralia (Barnden et al., 1995), the phenomenology of SSM (Davies, 1992;Probert, 1999), and the status of soft systems models (Gregory, 1993a)].

Second, the quality of understanding is mixed, regardless of the level ofdiscussion. There are accounts full of misunderstandings, accounts that lurchbetween understanding and misunderstanding through to sophisticated argument.

Third, recent reference to the early Lancaster work, work that inevitablyincludes more systems engineering characteristics than Checkland’s recent work,also complicates the picture. Here the observation that SSM developed at a pace“which might render consideration of the early work as anachronistic and dated”

Soft Systems Methodology: Other Voices 775

(Burrell, 1983, p. 125) remains apposite. Of interest is that Systems Thinking,Systems Practice (Checkland, 1981) is the most frequently cited primary source(although it is not always evident from the context which part of the book isbeing referred to). The obvious question in each particular case is, Is this deliber-ate or an oversight? Clare and Stuteley (1995), for example, which was published5 years after Soft Systems Methodology in Action cites only Systems Thinking,Systems Practice. Is this deliberate?

Finally, the secondary literature is becoming more influential in the sensethat reference to this literature for definition and justification, rather than to theprimary literature, is now occurring. For example, in “a root definition ‘is anidealised view of what a relevant system should be,’ ” Ellis (1995, p. 206) citesFlood and Jackson (1991, p. 197); Tudor and Tudor (1995) refer to descriptionsof Multiview (Avison and Wood-Harper, 1990) as the source of definitions; andBiggam and Hogarth (1996) cite work by Gregory, Jayaratna, Miles, Mingers,Ormrod, and Probert.

Differences between the work on SSM originating from Lancaster (Check-land and Wilson) have given rise to the notion of an SSM family or “genus”(Atkinson, 1986) of SSM methodologies; yielding, for example, the form of SSMin use in Australia (Barnden et al., 1995; Ledington and Donaldson, 1997) andin the United Kingdom (Mingers and Taylor, 1992) being described.

Differences have been noticed by other authors (e.g., Klein, 1991): in style[Checkland’s use is “liberal” and Wilson’s is “professional” (Atkinson, 1984,1986)] and in the form of root definitions and conceptual models (Forbes, 1989;Hitt, 1991). These have been seen as the inevitable consequences of use of amethodology rather than a method: differences that emerge from the interactionof methodology, the user of it, and the situation in which it is used.

This, together with Checkland’s “modern” SSM has resulted in referencesto the 1981 seven-stage version as the soft systems methodology (as an idealtype) (Atkinson, 1986), the “classical version” (Hirschheim et al., 1995), the“formal” (Atkinson, 1984), or the “generic version” (Watson and Smith, 1988).

2. UNDERPINNING ASSUMPTIONS

Holwell (1997) has argued that an interpretive foundation is a necessary(and a defining) characteristic of SSM. There is no equivalent argument to thisin the secondary literature, although the importance of the interpretive stance isnoticed in passing remarks, for example, “ssm relies on philosophical concepts insupport of its foundations” (Biggam and Hogarth, 1996, p. 249). The importanceof an interpretive perspective and/ or the centrality of subjectivity is evident inboth the basic and the more sophisticated accounts, although this is generally inthe context of a paraphrasing of Checkland’s work.

Detailed debate of the philosophical foundation of SSM is relatively limited,

Holwell776

although there are examples (including Atkinson, 1984; Flood and Carson, 1988;Rennie, 1989; Stowell, 1993; Visala, 1991). It is broadly agreed that it is withinthe phenomenology tradition (Atkinson, 1984; Crowe et al., 1996; Flood andCarson, 1988; Hirschheim et al., 1995; Jackson, 1982; Mingers, 1984; Probert,1991; Rennie, 1989); although it was not necessarily fully embraced in the earlywork (Burrell, 1983; Mingers, 1984; Rennie, 1989) or in the modeling (Daviesand Wood-Harper, 1989; Hirschheim et al., 1995). Others simply accept Check-land’s view that SSM is phenomenologically based and see this as a strength(Jayaratna and D’Arcy, 1993; Stowell and West, 1994; Visala, 1991; Willmott,1989).

The “interpretive” stance of SSM is more widely discussed and acceptedas a strength (Crowe et al., 1996; Dahlbom and Mathiassen, 1993; Doyle andWood, 1991; Flood and Jackson, 1991; Flood and Ulrich, 1991; Hitchins, 1991;Tsoukas, 1992; Walsham, 1993). Stowell (1993, p. 98) relates this to mode 2use especially, arguing that this is “the clearest example of interpretive systemsthat exists in a practical form”.

There is virtually no disagreement that SSM is an interpretive approach andthat this attribute is a particular strength of SSM. However, what is understoodby “interpretive”2 is varied—ranging from that it is based on the thinking of aparticular school of philosophy to simply that it takes subjectivity seriously (asin Lane, 1993).

Both the importance of subjectivity and that SSM takes it seriously fea-ture prominently in the literature (Avison, 1989; Benyon and Skidmore, 1987;Cavaleri, 1994; Crowe et al., 1996; Doyle and Wood, 1990; Lane, 1993; Mears-Young, 1995; Mingers, 1984; Rosenhead, 1989; Walsham, 1993). Similarly, but astep away from subjectivity as such, is the concern with meaning (Burrell, 1983;Feng, 1993; Romm, 1995) and the emphasis on understanding the problem-sit-uation in the fullest way (Avison and Wood-Harper, 1991; Harry, 1994; Masonand Willcocks, 1994; Wade, 1991; Wood and Doyle, 1989; Wood-Harper andFitzgerald, 1982). Associated with this are the points that SSM is a holisticapproach (Anderson, 1989; Jackson, 1982; Mingers and Taylor, 1992), i.e., thatit is not reductionist or “scientific” (Avison and Fitzgerald, 1988; Benyon, 1990;Lyytinen, 1987a; Rosenhead, 1989; Tudor and Tudor, 1995), and that its con-cern and outcome are learning (Brown, 1992; Lyytinen, 1988; Mathiassen et al.,1991) or sense making (Lyytinen, 1992).

Here the SSM concept of Weltanschauung and of multiple views of aproblem-situation are important (Avison et al., 1992; Cavaleri, 1994; Daviesand Ledington, 1991; Doyle and Wood, 1990; Lyytinen, 1988; Mathiassen etal., 1991; Mingers and Taylor, 1992; Romm, 1995; Skidmore, 1994; Walsham,

2 What is understood as constituting interpretivism varies considerably, and here an author’s use istaken at face value.


1993), as is that it is a pluralist (Flood and Jackson, 1991; Mason and Willcocks,1994) and/ or participative approach (Mason and Willcocks, 1994; Miles, 1988).

Three other features of SSM also link to an interpretive perspective. Firstis the use of “system” as a construct (or more recently as an epistemologicaldevice), i.e., systemicity in the process of inquiry rather than existing in theworld (Anderson, 1989; Cavaleri, 1994; Flood and Jackson, 1991; Mathiassenet al., 1991; Walsham, 1993), record the value of the conscious abstraction andreflection that comes from the above–below the line distinction (Benyon, 1990;Mathiassen et al., 1991; Saunders, 1989).

Finally, the consideration of social, political, and historical aspects rele-vant to a problem-situation are also important characteristics of SSM (Lyytinen,1988; Mason and Willcocks, 1994; Mingers and Taylor, 1992; Wood and Doyle,1989) [although the perceived social theory of SSM is seen by some as beinginadequate (Mansell, 1989; Willmott, 1989)]; indeed, Dudley (1995) sees cul-tural analysis being at the heart of SSM.

However, the perceived lack of subjectivity in the models (Jackson, 1991)or the view that the models are inherently functional (Hirschheim et al., 1995;Mingers, 1992) [albeit within the phenomenological SSM process (Davies andWood-Harper, 1989)] is noted.

So the interpretive foundation is seen as an integral part of Checkland’swork: an important and influential part (Rosenhead, 1989). This may be becauseof the particular philosophical foundation (e.g., phenomenology), or be lessspecifically defined as “interpretive” (however defined) through to the accep-tance of multiple meanings or the use of models as constructs in a learning pro-cess. Holwell’s (1997) argument that it is a necessary and defining part of SSMhas been adopted by Checkland (1999) in the retrospective.

It is this, the introduction of interpretive thinking (however it is understood)into the management and IS literature, that is a major contribution, if not thecontribution, to both the management and the IS fields by Checkland. There aretwo kinds of evidence for this. First, the extensive and widespread paraphras-ing of Checkland’s argument is acknowledgment of its significance; and second,there are examples of explicit recognition of this contribution. Flood and Jackson(1991), for example, explicitly acknowledge the contribution of

(a) using systems as an organizing framework for thinking, and not as arepresentation of reality; and

(b) identifying two “paradigms” (hard and soft), each with different assump-tions and therefore methodological principles, particularly the under-standing of social situations via perceptions of it.

They argue that Checkland’s distinction between “hard” and “soft” marks an“epistemological break” from positivism to interpretive systems thinking (Floodand Ulrich, 1991, p. 186). For others, Checkland “shook the systems move-

Holwell778

ment to its foundations by introducing the concepts and philosophy of inter-pretive social science into systems thinking” (Ellis, 1995, p. 212); thus mark-ing “a watershed” (Hopkins, 1995, p. 647) or “paradigm shift” (Crowe et al.,1996, p. 9). Others go further than this: Systems Thinking, Systems Practice hasbeen ranked as a pioneering work alongside the work of Bertalanffy and Simon(Yolles, 1994).

The hard–soft distinction is prominent in the literature, although, as used,the distinction has more to do with problem structure (Anderson, 1989; Combs,1995; Curtis, 1995; Harry, 1994; Tudor and Tudor, 1995) than with any shift ofsystemicity or paradigm.

Checkland’s work is recognized for adding interpretive thinking to the fieldsof systems, problem-solving, and IS: so much so, that his argument and languagehave become part of the general discourse.

The language and logic of Checkland’s work have been widely adopted intwo ways. First, the essence of the systems ideas is reproduced (in, e.g., Avisonand Wood-Harper, 1991; Rosenhead, 1989; Tudor and Tudor, 1995), includingthe systems typology (Fuenmayor, 1995), emergent property, adaptive system,systems hierarchy, communication, and control (Mathiassen et al., 1991), anddesirable and feasible changes (Benyon, 1990; Oakland, 1993; Skidmore, 1994).

Second, the arguments for looking at a problem-situation (rather than aproblem) (Avison et al., 1992; Curtis, 1995; Mason and Willcocks, 1994) and,therefore, consideration of the context as well as the perceived problem (Ander-son, 1989; Avison et al., 1992; Combs, 1995; Mason and Willcocks, 1994)are prominent. Such ill-defined problem-situations require a problem structuring(Skidmore, 1994) or problem definition (Mason and Willcocks, 1994) approachthat is not of the means–end kind (Avison et al., 1992; Tudor and Tudor, 1995),and one concerned with improvements, rather than finding a solution (Curtis,1995; Harry, 1994). In this context the cyclic, iterative nature of the SSM pro-cess also features prominently (Avison and Wood-Harper, 1991; Brown, 1992;Downs et al., 1992; Hirschheim and Schafer, 1988; Wade, 1991; Watson andSmith, 1988).

3. SSM AS PROCESS

The trend in the developments in Checkland’s work—the shift of systemic-ity, the increasing emphasis on “soft” notions (subjectivity, learning, multipleviews), and the move away from the seven-stage account to the logical andcultural two-stream account—does not feature in the secondary literature whendescribing the SSM as a process. This is sometimes contrary to the argumentfor SSM presented in the same account. The most recognized notions of SSM(as distinct from the whole body of Checkland’s work) are the rich picture [dia-gram (Lewis, 1992), rather than in the sense of a rich appreciation] and the root


definition/ conceptual model pairing. These techniques of SSM are described inevery overview account. This is evident from reading the literature and reflectsempirical findings (Ledington and Donaldson, 1997; Mingers and Taylor, 1992).The prominence of the rich picture is somewhat surprising—in one instancebeing elevated to definitive status in “the rich picture methodology of PeterCheckland” (Moyes, 1993, p. 144).

Accounts broadly reflect the state of Checkland’s work as set out in theearly 1980s; indeed many indicate (and others imply) that Systems Thinking,Systems Practice is the prime (and definitive) source. This is so even at the endof the 1990s. It is easy to find accounts of SSM published during the 1990sthat repeat the seven-stage description from Systems Thinking, Systems Practice(Bentley, 1992; Clare and Stuteley, 1995; Combs, 1995; Curtis, 1995; Oakland,1993; Wade, 1991). [Skidmore (1994) even reproduces a 1970 conceptual modelas an example.] So, the secondary literature accounts are of a seven-stage pro-cess conveyed in the language of hard versus soft (structured versus ill-defined)problems.

These seven-stage descriptions follow the original Constitutive Rules whichdefine several stages, each with a defined, required ouput. Moreover, the lan-guage used is also that which Checkland was using in the early 1980s: of hardversus soft (i.e., structured versus ill-defined) problems and approaches, prob-lem structuring, problem-content system, and found problem-solving system, etc.(in Avison and Fitzgerald, 1988; Clare and Stuteley, 1995; Curtis, 1995). Thislanguage has become part of the general management and IS discourse, in, forinstance, the notion of “soft OR” (Jackson, 1993; Rosenhead, 1989) and softmethods (Lehaney et al., 1997), but it does not include the “shift of systemic-ity” and places little or no emphasis on the nature of social reality.

So, fundamentally, the picture is little changed from that which Forbes(1989) found 10 years ago, namely, that the seven-stage process “defined” by theoriginal Constitutive Rules as described in Systems Thinking, Systems Practiceis the most common description of SSM.

The two-stream version has been noticed, although this is generally in thedeveloped literature (e.g., Avison and Fitzgerald, 1995; Crowe et al., 1996;Dahlbom and Mathiassen, 1993; Hirschheim et al., 1995; Stowell and West,1994). However, usually where the existence of both descriptions is noted, thetwo-stream “model” is set aside for pedagogical reasons and the authors revertto the seven-stage version, presumably for simplicity of exposition, although theinterpretive characteristics (the view of social reality, models as constructs, andAnalysis One, Two, Three) are acknowledged (e.g., in Davies and Ledington,1991; Harry, 1994; Hirschheim et al., 1995).

One development since Forbes’ (1989) review has significance. What hasbecome evident, particularly with the inclusion of SSM in IS textbooks, is dis-cussion of Checkland’s work in two distinctly separate parts: discussion of SSM

Holwell780

as a stark seven-stage process without elaboration or context, separated fromdiscussion of the balance of his work.

These accounts reproduce a comprehensive version of the main elementsof Checkland’s work, using the concepts, language, and logical structure ofit (Anderson, 1989; Avison and Fitzgerald, 1995; Combs, 1995; Curtis, 1995;Downs et al., 1992; Harry, 1994; Tudor and Tudor, 1995). However, this argu-ment may be separated from discussion of SSM by several chapters in a studenttext (as in Anderson, 1989; Benyon, 1990; Combs, 1995; Curtis, 1995; Harry,1994; Skidmore, 1994).

This division of Checkland’s work into two parts,

(1) human activity system modeling within a seven-stage framework (fre-quently described prescriptively) and

(2) a set of assumptions and notions relevant to the nature of problem-solv-ing in organizations,

is a significant development.

4. ERRORS, CONFUSION, AND MISUNDERSTANDING

In these accounts it is not uncommon for the argument to be used as ageneral framework of ideas and for the description of the seven-stage “SSM,”as either a technique or a modeling language, to be just one of several approachesor methods described later against the framework. The description of SSM insome accounts of this kind does not necessarily conform to the argument in thesame text: for example, the description of SSM that Checkland (1999, p. A42)cites (see below) shows little resemblance to the argument set out earlier in theoriginal text.

Furthermore, the extent to which “the argument” is attributed to Checklandvaries in these accounts. While his work is cited, this may only be at the level ofconcepts (such as the typology of systems), and where the whole, the form andcontent, is clearly a repetition of his work, this may not be acknowledged (as inCheung and Holden, 1993). Likewise, the concepts “human activity system” andWeltanschuuang, for example, are used freely and without acknowledgment, orby reference to the secondary literature. This is not the case with descriptionsof the stages of “SSM”; this is invariably acknowledged.

The decoupling of Checkland’s argument and process is significant in itself,but when associated with “incomplete” acknowledgment of the origins impliesthat Checkland’s work (argument) has become part of the taken-as-given foun-dation or conventional wisdom in both the management and the IS literatures.This is a significant achievement.

However, acceptance of the argument does not necessarily mean that it isunderstood. There are accounts that are fundamentally wrong; there are many


accounts containing inconsistencies and there are accounts which are reasonableexpositions of Checkland’s work, although these too frequently contain misun-derstandings.

The following, not untypical, and error-prone account that happens to comefrom an IS textbook illustrates some of the difficulties: What does it say aboutSSM? what does it convey about systems thinking? (Text set THUS INDICATES

ERRORS.)

This methodology stems from the work of Checkland (1981) who took a radi-cally different approach to the ANALYSIS AND DESIGN OF INFORMATIONHUMAN ACTIVITY SYSTEMS. Starting from the premise that ORGANIZATIONS(and therefore their SUBSYSTEM INFORMATION SYSTEMS) are open systemsthat INTERACT WITH THEIR ENVIRONMENT, he INCLUDES THE HUMANACTIVITY SUBSYSTEMS as part of his modelling process. The methodology startsby taking a particular view OF THE SYSTEM and incorporating subjective and objec-tive impressions into a ‘rich picture’ OF THE SYSTEM that includes the peopleinvolved, the problem areas, sources of conflict and other ‘soft’ aspects OF the overallsystem. A ‘root definition’ is then formed ABOUT THE SYSTEM which PROPOSESIMPROVEMENTS to THE SYSTEM to TACKLE THE PROBLEMS IDENTIFIEDIN THE RICH PICTURE.

Using THE root definition, VARIOUS conceptual models of the new systemcan be built, compared and evaluated against the PROBLEMS IN THE RICH PIC-TURE. A SET OF RECOMMENDATIONS is then suggested to deal with the specificchanges that are necessary to SOLVE THE PROBLEMS. These are evaluated in termsof feasibility and used to propose specific remedies for action.

This account could simply be dismissed as demonstrating an incompetentunderstanding of Checkland’s work (which it clearly is), but nevertheless, giventhat it is not untypical, it is relevant; after all, the authors considered “SSM”to be of sufficient importance to include it in their book, however they actuallyunderstand it.

Most errors are associated with the “existence,” or otherwise, of systemsand the nature of models. The most fundamental and common error is that sys-tems exist. This error can be found in most accounts, including those whichargue that systems are constructs, and are otherwise reasonable representationsof Checkland’s work: one, for example, states that models are epistemologicaldevices and then continues that a conceptual model is a model of the area ofinterest. This misunderstanding is also reflected in comments such as “identifythe correct system” and “SSM seeks to improve the human activity system.”Other examples are listed in Fig. 1.

Examples of careless reading of the material, such as that conceptual modelsare checked by the three E’s and that, because SSM is a holistic approach, itis not “hard,” are too easily found. An extreme example is the statement thatCheckland is the originator of critical systems thinking (Balle, 1994).

Analysis of the errors and misunderstandings is not relevant here, although

Holwell782

Fig. 1. Illustrative errors and misunderstandings from the secondary literature on SSM.

the sheer number of inadequate and/ or wrong accounts is important becausethe secondary literature is becoming more influential: not because it containsnew insights but because errors in one secondary source are being repeated inothers. Most commonly these are references to accounts of Multiview (whichuses SSM tools and elements of the argument) as being accounts of SSM useor are references to the descriptions of SSM by Flood and/ or Jackson. In otherwords, the primary literature is no longer necessarily the source of definitionsand argument.

The same trend is also evident in the more developed discourse. A paperthat is critical of aspects of SSM’s use of philosophical arguments (Biggam and


Hogarth, 1996) is based on accounts in the secondary literature, citing the sec-ondary literature in support without making adequate reference to Checkland’swork. Moreover, these references are to ideas of which Checkland would alsobe critical.

It is clear from the secondary literature that the nature of SSM is not under-stood, and not only because it is a relatively simple task to find flawed accountsof it. At the most basic level of what SSM “is” or what its purpose is, there arenot only different views, but clearly contradictory views.

At the most fundamental level: SSM is a general problem-solving approach(Mason and Willcocks, 1994; Wood-Harper and Fitzgerald, 1982) and SSM isnot a general problem-solving approach (Biggam and Hogarth, 1996; Jayaratna,1992). And from the IS literature, SSM is an information system developmentmethodology (Avison and Wood-Harper, 1991; Benyon, 1990; Hirschheim andSchafer, 1988; Watson et al., 1995) or it is not an information system devel-opment methodology (Downs et al., 1992; Harry, 1994; Mason and Willcocks,1994).

The purpose of enacting the SSM process is also confused. It is variously

• for organizational design or business analysis (Combs, 1995; Mingers andTaylor, 1992),

• for managing organizational change (Benyon, 1995), or• for general problem-solving (Gregory, 1993a; Mingers and Taylor, 1992;

Wood and Doyle, 1989; Wood-Harper and Fitzgerald, 1982), and• as an approach to learning (Hengerer, 1995; Lewis, 1994),• as an IS development approach (Avison and Wood-Harper, 1991; Lyyti-

nen, 1987b; Watson et al., 1995; Wood-Harper and Fitzgerald, 1982), or• as a front end for IS development (Avison and Fitzgerald, 1988;

Hirschheim and Schafer, 1988).

It is also understood as a research process (Brown, 1992); a loose framework, astate of mind (Hirschheim and Schafer, 1988); or a problem structuring approach(Goddard et al., 1994).

Whilst not all of these notions have foundations in the primary literature,their existence indicates a problem. A second development evident from exam-ination of the secondary literature has occurred without discussion and is iden-tified only with the benefit of hindsight.

There are two parts to it. First, a variety of terms has appeared in the liter-ature that would seem to flow from Checkland’s work and which often occur inconjunction with discussion of it. These are rarely described, let alone defined,but appear to be “common currency” and easily recognized, although uses ofthem show that they are variously understood. Thus you find reference to: softsystems, soft systems thinking, soft thinking, SSM, a soft systems methodology,SSM tools, soft systems practice, etc. (Figure 2 lists some additional examples.)

Holwell784

Fig. 2. Terms in the literature related to Checkland’s work.

Moreover, they are terms that have emerged out of the discourse of the inter-ested community, apparently unnoticed, reflecting a momentum of developmentoutside that described by the originators.

Some accounts are clear about the usage of terms in Fig. 2 but this is notnecessarily helpful in understanding the contribution of Checkland’s work. Forexample, Stowell (1995a, p. 126) uses the term “interpretive systems thinking,”adapting Checkland’s argument that the perceived world is problematic and thatsystemicity is in the process of inquiry, to define it. Stowell and West arguethat interpretive systems ideas is a philosophy of ideas compared to “the prac-tical (and often limited) use of Checkland’s Soft Systems Methodology (SSM)”(1994, p. 111). While it is all too easy to find examples of “limited” use andunderstanding of SSM, it does seem unnecessary to invent a new term for aframework of ideas that has already been set out.

But elsewhere it is soft systems thinking that is defined in terms of these twolinked assumptions—use of systems as constructs and perceived social reality(Dahlbom and Mathiassen, 1993). In other words, the same two assumptionsdefine interpretive systems thinking and soft systems thinking: suggesting thatthey are the same thing. For others, soft systems thinking is simply systemsthinking that recognises subjectivity (Gammack, 1995).

From both the emergence of the terms listed in Fig. 2 and the explicit linkingof the two assumptions relating to social reality and systemicity, it would seemthat there is a change in the understanding of “soft” in “soft systems methodol-ogy” from the early sense of structure to a meaning that is linked to either thesetwo assumptions or subjectivity.


5. OTHER VOICES ON IS WORK

The three aspects of the IS argument most noticed in the IS-related sec-ondary literature are

(1) the importance of meaning (including multiple views of meaning),(2) that IS and IS development are concerned with more than technology,

and(3) the idea of understanding the “system served” before considering IS

design.

These notions reflect the concerns of those people interested in an interpre-tive, rather than an objectivist, approach to IS.

Just as with the general SSM material, the acknowledged contribution ofCheckland’s work comes from the introduction of interpretive ideas into ISthinking. Again, the accreditation varies from the specifics of phenomenology(Hirschheim et al., 1995) through interpretivism (Walsham, 1993) generally, par-ticularly as an operationalization of interpretive ideas (Dahlbom and Mathiassen,1993; Lewis, 1994; Stowell and West, 1994) to that it takes subjectivity (mean-ing) seriously.

There are some problems with the “empirically” derived list (Fig. 3) whichmerely records what has been noticed. Whether all of these characteristics arejustifiably associated with the Lancaster IS work is debatable. Two examplesillustrate this: first, the implication that SSM is an IS development approachmentioned earlier—Lyytinen (1987a), for example, makes this point but is ques-tionable [particularly as the reference cited is Checkland (1981)]; and, second,that use of the approach “forces user involvement” (Benyon, 1995) in the pro-cess of IS development is not sustainable against any primary account of theLancaster work.

A significant problem encountered when undertaking this review is implicitin these two examples—both what is taken to be the Lancaster work relevant to ISis confused and the quality of understanding of it is so mixed as to make critical useof it difficult. There are references to Checkland (1981) in IS work as being aboutSSM and/ or IS; there are references to specific parts of Checkland’s work (e.g.,Checkland and Scholes, 1990, pp. 53–58) that are clearly relevant to IS, and thereare isolated references to Wilson’s IS work: and all of these are likely to attach thelabel “SSM” to what they perceive and describe. It may refer to Checkland’s pre-1990 seven-stage SSM, Checkland’s two-stream SSM, Wilson’s human activitysystem modeling, Wilson’s (1990) information systems methodology, or Check-land’s IS work (1984, 1988; Checkland and Holwell, 1993).

The general “impressions” conveyed in the secondary literature are rela-tively straightforward at one level and difficult to take seriously on another. Thefirst of the straightforward characteristics is the separation of the argument from

Holwell786

Fig. 3. SSM characteristics noticed as relevant to IS.

description of the “SSM” process that has already been mentioned. [Use of “theargument” is decoupled from the “SSM” process in a different way in the bookby Stowell and West (1994) on Client-Led Design. There, the argument has beenadopted by the authors and they distinguish their work from “mechanical” useof “SSM.”]


In such accounts the language is that of the 1980s, which is not surprisinggiven that most accounts refer to Checkland (1981) rather than to Checkland andScholes (1990). This is the language of hard and soft problems, problem contentand problem-solving systems, and fundamental systems concepts rather than aninterpreted social reality and the “shift of systemicity.”

Other accounts indirectly complicate the overall picture. The inclusion ofSSM as one of six IS methodologies in a book titled Information Systems Devel-opment and Data Modeling (Hirschheim et al., 1995) identifies it as an ISDapproach, and yet it is “classical” SSM that is described in this book. This istypical of the confusion in the secondary literature. Quite how “classical” SSMis an IS development methodology, as this is generally understood, is not dis-cussed.

Similarly, Avison and Fitzgerald (1995, p. 11) set out six objectives for anIS methodology: that it record requirements accurately, is a systematic processof development so progress can be monitored, produces an acceptable IS for anacceptable cost in an acceptable time, produces a well-documented and easy-to-maintain IS, identifies changes early, and produces an IS liked by users. Despitehaving these objectives, an account of SSM is included without comment orreference to Checkland’s IS process or argument. How the account of classic“SSM” that is included relates to these objectives is not discussed, and yet theinclusion of “SSM” in the book is not accidental. This mismatch is not com-mented on, just as the gap between the characteristics for which Checkland’swork is noticed and the impoverished accounts of “SSM” is not mentioned byothers.

The argument for use of SSM in IS, then, is likely to be based on theimportance of understanding the context, and so the relevance of using SSMderives from the view that IS design is both a “soft” problem and a problemthat is concerned with meaning. In the developed discourse, on the other hand,the argument is generally based on notions of an interpreted social reality thatis known via mental models. However, use of the term “the argument” in bothdiscourses is misleading because both the general argument (i.e., for SSM) andthe IS argument are used in different accounts. Moreover, the difference is notcommented on. Unexpectedly, the general argument is more prominent in IStextbooks accounts than the IS argument.

So, the ideas have been influential in IS thinking: as evidenced by, forexample, the existence of a set of books on IS influenced by Checkland’s work(Dahlbom and Mathiassen, 1993; Davies and Ledington, 1991; Jayaratna, 1994;Stowell, 1995b; Stowell and West, 1994), that the arguments have become partof the framework of ideas deemed necessary for students of IS, and have beenexplicitly acknowledged as an important contribution to the field—primarilybecause they are perceived as being part of the interpretive strand of thinking.

The actuality of the process described to realize the argument, however, is

Holwell788

somewhat different. There is an interpretivist argument espoused and the char-acteristics of “SSM” that have been noticed are those most clearly associatedwith such an argument (multiple views, learning, iteration, models relevant to,etc.), but the dominant process descriptions are of a “prescriptive” or mechanicalversion of the seven-stage process. These accounts generally follow the originalConstitutive Rules and describe each of the seven stages of “classical” SSM andtheir “products.” Crucially, stages 3 and 4 (root definition and conceptual model)are emphasized.

This emphasis on modeling from a root definition and the perception of arich picture (as a product rather than an appreciation of a situation) frequentlygives the impression that these three techniques not only are useful tools but thatthey constitute the essence of the process (and the SSM approach).

The two-stream version is rarely noticed in the IS literature, and where itis noticed it is set aside, sometimes deliberately because the “classic” version isseen as being more applicable to IS (Harry, 1994; Hirschheim et al., 1995). (Whythis would be so, given, for example, the emphasis on social considerations inthe interpretive IS argument, is not clear.)

The developed discourse is more concerned with the assumptions andconcepts, whereas modeling is the prime focus of the less sophisticated dis-course, although the interpretive argument and characteristics are recognized.This emphasis on modeling, together with the decoupling of the argument andprocess, results in impoverished versions of the seven stages of SSM whichunduly emphasize modeling at the expense of process—although they may havebeen preceded by expositions of Checkland’s argument about problem-solvingin organizations.

This is not dissimilar to Checkland’s own work in that an interpretiveargument is associated with the sparse process (described by Checkland andScholes, 1990, p. 57), but there is an important difference. The secondary liter-ature accounts are of SSM as a general problem-solving methodology. There isno link to IS analysis and design from such a process, and surprisingly in mosttextbook accounts this crucial point is not raised. The relevance of the classicSSM to IS in general or to IS development is either left to the imagination ofthe reader (because nothing is said) or an adaptation is attempted.

Beyond this level of observation, use of the secondary literature becomesmore difficult. Reference to the secondary rather than the primary literature forconcept definition has already been noted, but this is particularly important ifthe source referred to is a Multiview description. Multiview makes use of sometechniques from SSM, such as rich picture and human activity system models,but with a different purpose and status to their use in SSM. In Multiview ahuman activity system model built from a rich picture is the first of five stagesof analysis. The authors of Multiview recognize that their use of these techniquesis different from that of Checkland; others are not so careful. It is clear that other


authors are not aware of this and unknowingly repeat the Multiview use as ifthis were equivalent to use in SSM. This trend has the effect of adding to theimpression that this techniques “constitute” the “SSM” approach noted earlier,and this in turn adds to (and reflects) the misunderstanding of Checkland’s work.

Summarizing then, Checkland’s work, both the general and the IS work,is noticed because of its interpretivist foundation and characteristics; and boththe general argument and the IS argument are repeated (i.e., acknowledged) andrecognized as being influential in IS thinking. Even though the ideas are influ-ential, Checkland’s IS process (Checkland and Scholes, 1990, p. 57) does notfeature in the literature. The “omission” is passed over without debate. At thesame time, techniques, particularly rich picture, root definition, and conceptualmodel, have also been adopted and used independently of Checkland’s work.

Two issues further complicate the picture: first, there is the question ofwhether the concern of the work is with exploration of the context, or require-ments, or both; and, second, there are contradictory views of “SSM.” Some view“SSM” (usually a sparse classical SSM) as an ISD methodology (Avison andWood-Harper, 1991; Hirschheim and Schafer, 1988; Lyytinen, 1987a; Watsonet al., 1995) and may even cite Systems Thinking, Systems Practice (Checkland,1981): although neither this statement, nor reference, is justifiable against theprimary literature. Others, rightly, point out that it is not an ISD methodologyand may propose a role for “SSM” as either a front end to some more traditionalapproach or propose that it be the overarching framework for the developmentprocess.

5.1. Adaptations

One indication of the influence of the ideas is the number of extensionsand adaptations that have been proposed. Of these, both Multiview and Client-Led Design have been mentioned. Others include Functional Analysis for OfficeRequirements (FAOR) (Hirschheim, 1985; Schafer, 1988) and Saywer’s OPIUM(1992). Lewis (1994) has taken a different route.

What these different methods have in common is that they are attemptsto link the learning and understanding of the situation and requirements gainedfrom an “SSM” inquiry into the requirements of developing a real-world IS.They do this by making use of either techniques or both the ideas and sometechniques (especially human activity system modeling).

Slightly differently, Checkland’s work has been linked to IS design at thelevel of trying to link SSM techniques, primarily conceptual models, to tradi-tional IS techniques. Benyon and Skidmore (1987) proposed linking concep-tual models to data flow diagrams based on the perceived similarities of them.Mingers (1988), in reply, acknowledged some similarities but indicated that therewere fundamental philosophical differences between them; Avison et al. (1998)

Holwell790

also point out the philosophical difference and conclude that they are not simi-lar. Other proposals include linking “SSM” to the Jackson System Developmentmethod (Savage and Mingers, 1996; Wood and Doyle, 1989) and re-expressinga model in a form that can be expressed in propositional calculus and predicatelogic in order to derive a type of DFD (Gregory, 1993a–c; Merali, 1992).

A third facet relates to the argument that different philosophical assump-tions underpin “SSM” and traditional IS techniques. Discussion of this is atthe center of the “grafting vs. embedding” debate. Miles (1988, 1992) identi-fied two ways of using the ideas in IS: grafting on as a front end to conven-tional approaches, and embedding other IS techniques within an overall SSMprocess. For a time the focus of this debate came to rest largely on the questionof the incommensurability of the paradigms presumed for SSM (interpretive)and IS (objectivist). [Contributions include Beeby (1993), Doyle et al., (1993),Doyle and Wood (1991), Miles (1992), Prior (1992), Wood (1992), and Woodand Doyle (1989).]

Others point out that while such questions may be of importance to aca-demics, it either does not cause problems in practice or may be overcome(Sawyer, 1992; Stowell and West, 1994). Doyle et al. (1993) suggest a rationalebased on a distinction between soft and hard as being concerned with “doing theright thing” and “doing the thing right,” respectively; and it is discussed prag-matically in Crowe et al. (1996). In more recent years this debate has “evapo-rated” and attention turned elsewhere, although it remains essentially unresolved.

Thus far three options for using the Lancaster ideas in IS have been iden-tified (Jones, 1992). These are, first, use as a stand-alone method; second, useas part of an eclectic methodology (such as in Multiview) and as a basis forstructured methods (such as in FAOR and OPIUM)—the most common is touse it as a front end to traditional IS methods, which is clearly the case withthe addition of a mechanistic SSM to the Feasibility Stage of SSADM (CCTA,1993); and, third, use of SSM as an overarching framework for the ISD pro-cess. The developers of Client-led Design would probably not see it as being anexample of this approach because they see it as being a “philosophy” or frame-work of ideas, whereas SSM is a methodology. Nevertheless, the ideas they usehave come to IS via Checkland’s work, as have the tools, techniques, and con-cepts used in the early phases. Moreover, the delineation of their work fromimpoverished versions of seven-stage “SSM” has been through the adoption ofthe framework of assumptions that Holwell (1997) argues is a necessary part ofCheckland’s SSM.

6. THE NATURE OF SSM

Differences between the accounts of SSM by Checkland and Wilson (1990)have been described in terms of three categories:


(1) assumptions,(2) process, and(3) concepts and language (Holwell, 1996).

Topics in the SSM secondary literature can also be grouped into three categories(Holwell, 1997):

(1) argument, concepts, and assumptions;(2) the seven-stage prescriptive process; and(3) techniques (modeling and rich picture).

Others have also used three categories or levels to describe aspects of SSM,for example, modes of use (Barnden et al., 1995) and SSM practice (Jayaratna,1996),3 and Ledington (1989) identified “three levels of abstraction” in the dis-course about SSM. Although there are no well-developed arguments in the lit-erature for these schema, they have an important feature in common. In all ofthem the philosophical assumptions are identified as a separate category.

Ledington and Donaldson (1997) go further, stating that conscious adop-tion of a particular stance is required to substantiate a claim to be using SSM.Holwell (1997) sets out an argument for making this a necessary and definingcharacteristic of SSM.

This view is in marked contrast to the secondary literature as a whole. Thereis no discussion or understanding of the relationship between the two differ-ent elements of Checkland’s work: the argument (underpinning philosophy andassumptions) and process (SSM). An exception to this is Avison and Fitzgerald(1995), although their link is novel. They assume that the F, M, A concepts arepart of SSM and that the interpretive argument is part of “F” and so is appliedvia the SSM “M.” Checkland has not argued that this is a part of SSM, butit is nevertheless a not unreasonable view. This lack of argument is crucial. Itmeans that any relationship relies on either assumption or an association thatmerely rests on both elements being described in a particular contribution. Thisis broadly the case with Checkland’s own work.

7. CONCLUSION

The difficulties in making use of the secondary literature are inherent in thepoints made so far. First, it is not obvious what is meant by “SSM”; second, itis not clear how much (if any) of the ideas framework is assumed to be associ-ated with the process; third, the emphasis on modeling techniques reduces theemphasis on process; finally, and crucially, there is the generally poor under-standing of the work. These points relate to “SSM” as a general methodologyin the IS literature, and to this are added the difficulties that arise because the3 At the University of Central Lancashire, March 1996.

Holwell792

“role” of “SSM” in IS as set out in the literature is confused. All of these fac-tors add to the confusion surrounding the work and also demonstrate a lack ofunderstanding of it.

Checkland’s work as a whole is not well understood; this is despite it beingrecognized as an important contribution. Soft systems thinking is perceived asbeing “different” to SSM (Benyon, 1990; Lyytinen, 1987a; Wood-Harper andFitzgerald, 1982) and some techniques are being used explicitly outside SSM(Avison and Fitzgerald, 1995).

A set of ideas, essentially from Checkland’s work, is acknowledged as beingimportant (important enough to be included in student texts), but at the same timeit seems that in many instances it is difficult to see why, given the accompanyingaccounts of “SSM.” For anyone interested in critical use of the ideas, then thesecontradictions, gaps, and poor understanding make general use of the secondaryliterature something of a dubious undertaking.

REFERENCES

Anderson, R. (1989). Development of Business Information Systems, Blackwell, Oxford.Atkinson, C. (1984). Metaphor and Systemic Praxis, Unpublished Ph.D. thesis, Lancaster University,

Lancaster.Atkinson, C. J. (1986). Towards a plurality of soft systems methodology. J. Appl. Syst. Anal. 13,

19–31.Avison, D. E. (1989). An overview of information systems development methodologies. In Flood,

R. L., Jackson, M. C., and Keys, P. (eds.), Systems Prospects: The Next Ten Years of SystemsResearch, Plenum, New York, pp. 189–193.

Avison, D. E., and Fitzgerald, G. (1988). Information Systems Development: Methodologies, Tech-niques and Tools, Blackwell, Oxford.

Avison, D. E., and Fitzgerald, G. (1995). Information Systems Development: Methodologies, Tech-niques and Tools, 2nd ed., McGraw–Hill, London.

Avison, D. E., and Wood-Harper, A. T. (1990). Multiview: An Exploration in Information SystemsDevelopment, Alfred Waller, Henley-on-Thames.

Avison, D. E., and Wood-Harper, A. T. (1991). Information systems development research: An explo-ration of ideas in practice. Comput. J. 34(2), 98–112.

Avison, D. E., Fitzgerald, G., and Wood-Harper, A. T. (1988). Information systems development: Atoolkit is not enough. Comput. J. 31(4), 379–380.

Avison, D. E., Golder, P. A., and Shah, H. U. (1992). Towards an SSM toolkit: Rich picture dia-gramming. Eur. J. Inform. Syst. 1(6), 397–407.

Balle, M. (1994). Managing with Systems Thinking: Making Dynamics Work for You in BusinessDecision Making, McGraw–Hill, London.

Barnden, A., Smith, R., and Watson, R. B. (1995). A framework for analysing SSM-based studies. InEllis, K., Gregory, A., Mears-Young, B. R., and Ragsdell, G. (eds.), Critical Issues in SystemsTheory and Practice, Plenum, New York, pp. 207–213.

Beeby, R. B. (1993). On the incommensurability of hard and soft systems approaches to informa-tion system provision. In Stowell, F. A., West, D., and Howell, J. G. (eds.), Systems Science:Addressing Global Issues, Plenum, New York, pp. 307–312.

Bentley, T. (1992). Soft systems methodology. Manage. Account. July/ August, 22.


Benyon, D. (1990). Information and Data Modelling, Blackwell, Oxford.Benyon, D. (1995). Information and Data Modelling, 2nd ed., Blackwell, Oxford.Benyon, D., and Skidmore, S. (1987). Towards a toolkit for the systems analyst. Comput. J. 30(1),

2–8.Biggam, J., and Hogarth, A. (1996). Soft systems methodology (SSM): Fundamental issues. In

Jayaratna, N., and Fitzgerald, B. (eds.), Lessons Learned from the Use of Methodologies, BritishComputer Society, Information Systems Methodologies Specialist Interest Group, pp. 249–255.

Brown, A. D. (1992). Grounding soft systems research. Eur. J. Inform. Syst. 1(6), 387–395.Burrell, G. (1983). ‘Systems Thinking, Systems Practice’: A review. J. Appl. Syst. Anal. 10, 121–126.Cavaleri, S. A. (1994). ‘Soft’ systems thinking: A pre-condition for organizational learning. Hum.

Syst. Manage. 13, 259–267.CCTA (1993). Applying Soft Systems Methodology to an SSADM Feasibility Study, HMSO, London.Checkland, P. B. (1981). Systems Thinking, Systems Practice, John Wiley & Sons, Chichester.Checkland, P. B. (1984). Systems theory and information systems. In Bemelmans, T. M. A. (ed.),

Beyond Productivity: Information System Development for Organizational Effectiveness, North-Holland, Amsterdam, pp. 9–21.

Checkland, P. B. (1988). Information systems and systems thinking: Time to unite? Int. J. Inform.Manage. 8(4), 239–248.

Checkland, P. B. (1999). Soft Systems Methodology in Action: Including a 30 Year Retrospective,John Wiley & Sons, Chichester.

Checkland, P. B., and Holwell, S. E. (1993). Information management and organizational processes:An approach through soft systems methodology. J. Inform. Syst. 3, 1–15.

Checkland, P. B., and Scholes, J. (1990). Soft Systems Methodology in Action, John Wiley & Sons,Chichester.

Cheung, L. C-Y., and Holden, T. (1993). Information requirements. In Stowell, F. A., West, D.,and Howell, J. G. (eds.), Systems Science: Addressing Global Issues, Plenum, New York, pp.319–324.

Clare, C., and Stuteley, G. (1995). Information Systems—Strategy to Design, Chapman Hall, London.Combs, M. R. (1995). Information Systems for Business Management, Pitman, London.Crowe, M., Beeby, R., and Gammack, J. (1996). Constructing Systems and Information: A Process

View, McGraw–Hill, London.Curtis, G. (1995). Business Information Systems: Analysis, Design and Practice, 2nd ed.,

Addison–Wesley, Wokingham.Dahlbom, B., and Mathiassen, L. (1993). Computers in Context: The Philosophy and Practice of

Systems Design, Blackwell, Oxford.Davies, L. J. (1992). On the reality of systems. Syst. Pract. 5(6), 601–627.Davies, L. J., and Ledington, P. (1991). Information in Action, Macmillan, London.Davies, L. J., and Wood-Harper, A. T. (1989). Information system development: Theoretical frame-

work. J. Appl. Syst. Anal. 16, 61–73.Downs, E., Clare, P., and Coe, I. (1992). Structured Systems Analysis and Design Method: Applica-

tion and Context, 2nd ed., Prentice Hall, Hemel Hempsted.Doyle, K. G., and Wood, R. G. (1990). Soft Systems and Systems Enginering: Across the Great

Divide, Bristol Polytechnic, Bristol, UK.Doyle, K. G., and Wood, R. G. (1991). Systems thinking, systems practice: Dangerous liaisons.

Systemist 13(1), 28–30.Doyle, K. G., Wood, J. R. G., and Wood-Harper, A. T. (1993). Soft systems and systems engineer-

ing: On the use of conceptual models in information system development. J. Inform. Syst. 3,187–198.

Dudley, D. (1995). Surviving seeing medusa. In Ellis, K., Gregory, A., Mears-Young, B. R., and

Holwell794

Ragsdell, G. (eds.), Critical Issues in Systems Theory and Practice, Plenum, New York, pp.459–463.

Ellis, R. K. (1995). Critical considerations in the development of systems thinking and practice. Syst.Pract. 8(2), 199–214.

Fairtlough, G. (1982). A note on the use of the term ‘weltanschauung’ (W) in Checkland’s “SystemsThinking, Systems Practice.” J. Appl. Syst. Anal. 9, 131–132.

Feng, J-K. (1993). Conceptual modelling and DBS/ KBS Design. In Stowell, F. A., West, D., andHowell, J. G. (eds.), Systems Science: Addressing Global Issues, Plenum, New York, pp.337–342.

Flood, R. L., and Carson, E. R. (1988). Dealing with Complexity: An Introduction to the Theoryand Application of Systems Science, Plenum, New York.

Flood, R. L., and Jackson, M. C. (1991). Creative Problem Solving: Total Systems Intervention, JohnWiley & Sons, Chichester.

Flood, R. L., and Ulrich, W. (1991). Testament to conversations on critical systems thinking betweentwo systems practitioners. In Flood, R. L., and Jackson, M. C. (eds.), Critical Systems Thinking:Directed Readings, John Wiley & Sons, Chichester, pp. 185–206.

Forbes, P. (1989). The Development and Dissemination of Soft Systems Methodology, UnpublishedPh.D. thesis, Lancaster University, Lancaster.

Fuenmayor, R. (1995). The will to systems. In Ellis, K., Gregory, A., Mears-Young, B. R., andRagsdell, G. (eds.), Critical Issues in Systems Theory and Practice, Plenum, New York, pp.25–33.

Gammack, J. (1995). Modelling subjective requirements objectively. In Stowell, F. (ed.), InformationSystems Provision: The Contribution of Soft Systems Methodology, McGraw–Hill, London, pp.159–185.

Goddard, A., Connell, C., Philp, I., and Bray, J. (1994). Stakeholder participation approach andinformation system development: An action research case study, Discussion Paper No. 94-83,Department of Accounting & Management Science, University of Southampton, Southampton.

Gregory, F. H. (1993a). Cause, effect, efficiency, and soft systems models. J. Operat. Res. Soc. 44(4),333–344.

Gregory, F. H. (1993b). Logic and meaning in conceptual models: Implications for information sys-tem design. Systemist 15(1), 28–43.

Gregory, F. H. (1993c). Soft systems methodology to information systems: A Wittgensteinianapproach. J. Inform. Syst. 3, 149–168.

Harry, M. (1994). Information Systems in Business, Pitman, London.Hengerer, O. (1995). Strategic Considerations on Soft Systems Methodology, Unpublished M.Sc.

thesis, Lancaster University, Lancaster.Hirschheim, R. (1985). Office Automation: A Social and Organizational Perspective, John Wiley &

Sons, Chichester.Hirschheim, R., and Schafer, G. (1988). Requirements analysis: A new look at an old topic. J. Appl.

Syst. Anal. 15, 101–118.Hirschheim, R., Klein, H., and Lyytinen, K. (1995). Information Systems Development and Data

Modeling: Conceptual and Philosophical Foundations, Cambridge University Press, Cam-bridge.

Hitchins, D. K. (1991). A unified systems hypothesis. In Jackson, M. C., Mansell, G. J., Flood,R. L., Blackham, R. B., and Probert, S. V. E. (eds.), Systems Thinking in Europe, Plenum, NewYork, pp. 207–214.

Hitt, R. (1991). A Quantitative Investigation of SSM Modelling Techniques, Unpublished M.Sc. the-sis, Lancaster University, Lancaster.

Holwell, S. E. (1996). SSM: Theory and Practice. Proceedings of SSM Research Discussion Work-shop, Preston, 15 March.


Holwell, S. E. (1997). Soft Systems Methodology and Its Role in Information Systems, UnpublishedPh.D. thesis, Lancaster University, Lancaster.

Hopkins, B. (1995). Critical thinking in the information systems undergraduate curriculum. In Ellis,K., Gregory, A., Mears-Young, B. R., and Ragsdell, G. (eds.), Critical Issues in Systems Theoryand Practice, Plenum, New York, pp. 647–651.

Jackson, M. C. (1982). The nature of ‘soft’ systems thinking: The work of Churchman, Ackoff andCheckland. J. Appl. Syst. Anal. 9, 17–29.

Jackson, M. C. (1991). The origins and nature of critical systems thinking. Syst. Pract. 4(2), 131–149.Jackson, M. C. (1993). Social theory and operational research practice. J. Operat. Res. Soc. 44(6),

563–577.Jayaratna, N. (1992). Should we link SSM with information systems. Systemist 14(3), 108–119.Jayaratna, N. (1994). Understanding and Evaluating Methodologies, McGraw–Hill, London.Jayaratna, N., and D’Arcy, B. D. (1993). Desirability and feasibility of linking ‘soft’ systems method-

ology (SSM) with structured methodologies. In Stowell, F. A., West, D., and Howell, J. G.(eds.), Systems Science: Addressing Global Issues, Plenum, New York, pp. 637–642.

Jones, M. (1992). SSM and information systems. Systemist 14(3), 123–125.Klein, J. K. (1991). Systems: Concepts, Methodologies and Applications (second edition). Book

review. Syst. Pract. 4(5), 528–530.Lane, D. C. (1993). With a little help from our friends: How third generation system dynamics and

the issue structuring techniques of ‘soft’ OR can learn from each other, CUBS Working PaperNo. ITM/ 93/ CDL2, Information Technology Division, City University Management School,London.

Ledington, P. (1989). An issue-based analysis of the development of soft systems methodology. InFlood, R. L., Jackson, M. C., and Keys, P. (eds.), Systems Prospects: The Next Ten Years ofSystems Research, Plenum, New York, pp. 257–260.

Ledington, P., and Donaldson, J. (1997). Soft OR and management practice: A study of the adoptionand use of soft systems methodology. J. Operat. Res. Soc. 48(3), 229–240.

Lehaney, B., Martin, S., and Clarke, S. (1997). A review of problem structuring methodologies.Systemist 19(1), 11–28.

Lewis, P. J. (1992). Rich picture building in the soft systems methodology. Eur. J. Inform. Syst. 1(5),351–360.

Lewis, P. J. (1994). Information-Systems Development, Pitman, London.Lyytinen, K. (1987a). Different perspectives on information systems: Problems and solutions. ACM

Comput. Surv. 19(1), 5–46.Lyytinen, K. (1987b). A taxonomic perspective of information systems development. In Boland,

R. J., and Hirschheim, R. A. (eds.), Critical Issues in Information Systems Research, John Wiley& Sons, Chichester, pp. 3–41.

Lyytinen, K. (1988). Stakeholders, information system failures & soft systems methodology: Anassessment. J. Appl. Syst. Anal. 15, 61–81.

Lyytinen, K. (1992). Information systems and critical theory. In Alvesson, M., and Willmot, H.(eds.), Critical Management Studies, Sage, London, pp. 159–180.

Mansell, G. (1989). Dualism in social theory and systems practice. In Jackson, M. C., Keys, P., andCropper, S. A. (eds.), Operations Research and the Social Sciences, Plenum, New York, pp.171–175.

Mason, D., and Willcocks, L. (1994). Systems Analysis, Systems Design, Alfred Waller, Henley-on-Thames.

Mathiassen, L., Munk-Madsen, A., Nielsen, P. A., and Stage, J. (1991). Soft systems in softwaredesign. In Jackson, M. C., Mansell, G. J., Flood, R. L., Blackham, R. B., and Probert, S. V. E.(eds.), Systems Thinking in Europe, Plenum Press, New York, pp. 311–317.

Mears-Young, B. (1995). Logistics awareness. In Ellis, K., Gregory, A., Mears-Young, B. R., and

Holwell796

Ragsdell, G. (eds.), Critical Issues in Systems Theory and Practice, Plenum, New York, pp.577–583.

Merali, Y. (1992). Analytic data flow diagrams: An alternative to physicalism. Systemist 14(3),190–198.

Miles, R. K. (1988). Combining ‘soft’ and ‘hard’ systems practice: Grafting or embedding? J. Appl.Syst. Anal. 15, 55–60.

Miles, R. K. (1992). Combining ‘hard’ and ‘soft’ systems practice: Grafting and embedding revisited.Systemist 14(2), 62–66.

Mingers, J. (1984). Subjectivism and soft systems methodology—A critique. J. Appl. Syst. Anal. 11,85–103.

Mingers, J. (1988). Comparing conceptual models and data flow diagrams. Comput. J. 31(4),376–379.

Mingers, J. (1992). Questions and suggestions in using soft systems methodology. Systemist 14(2),54–61.

Mingers, J., and Taylor, S. (1992). The use of soft systems methodology in practice. J. Operat. Res.Soc. 43(4), 321–332.

Moyes, P. (1993). Using pictures to plan and manage change. Systemist 15(3), 141–147.Oakland, J. S. (1993). Total Quality Management, 2nd ed., Butterworth Heinemann.Pidd, M. (1996). Tools for Thinking: Modelling in Management Science, John Wiley & Sons, Chich-

ester.Prior, R. (1992). Linking SSM and IS development. Systemist 14(3), 128–132.Probert, S. K. (1991). A Critical Study of the National Computing Centre’s Systems Analysis and

Design Methodology and Soft Systems Methodology, Unpublished M. Phil thesis, Newcastleupon Tyne Polytechnic, Newcastle.

Probert, S. K. (1999). A critical analysis of the practical relationship between soft systems method-ology, Husserlian phenomenology and systems development. In Wood-Harper, A. T., Jayaratna,N., and Wood, J. R. G. (eds.), Methodologies for Developing and Managing Emerging Tech-nology Based Information Systems, Springer-Verlag, London, pp. 15–23.

Rennie, H. G. (1989). North Labrador and the Torngat Co-Op: An Exploration of Checkland’s SoftSystems Methodology Through Its Application to Fisheries Development, Unpublished M.A.thesis, Memorial University of Newfoundland, Newfoundland.

Romm, N. (1995). Knowing as intervention: Reflections on the application of systems ideas. Syst.Pract. 8(2), 137–167.

Rosenhead, J. (1989). Diverse unity: The principles and prospects for problem structuring meth-ods. In Rosenhead, J. (ed.), Rational Analysis for a Problematic World, John Wiley & Sons,Chichester, pp. 341–358.

Saunders, B. (1989). Be prepared—With soft analysis. OR Insight 2(2), 11–14.Savage, A., and Mingers, J. (1996). A framework for linking soft systems methodology (SSM) and

Jackson System Development (JSD). Inform. Syst. J. 6(2), 109–131.Sawyer, K. (1992). A contribution towards the debate on linking SSM to IS. Systemist 14(3),

199–201.Schafer, G. (1988). Functional Analysis of Office Requirements: A Multiperspective Approach,

J. Wiley & Sons, Chichester.Skidmore, S. (1994). Introducing Systems Analysis, Blackwell, Oxford.Stowell, F. (1993). Hermeneutics and organisational inquiry. Systemist 15(2), 87–100.Stowell, F. (1995a). Empowering the client: The relevance of SSM and interpretivism to Client-Led

Design. In Stowell, F. (ed.), Information Systems Provision: The Contribution of Soft SystemsMethodology, McGraw–Hill, London, pp. 118–139.

Stowell, F. (ed.) (1995b). Information Systems Provision: The Contribution of Soft Systems Method-ology, McGraw–Hill Information Systems, Management and Strategy, McGraw–Hill, London.


Stowell, F., and West, D. (1994). Client-Led Design: A Systemic Approach to Information SystemsDefinition, McGraw–Hill, Maidenhead.

Tsoukas, H. (1992). Panoptic reason and the search for totality: A critical assessment of the criticalsystems perspective. Hum. Relat. 45(7), 637–657.

Tudor, D. J., and Tudor, I. J. (1995). Systems Analysis and Design: A Comparison of StructuredMethods, NCC Blackwell, Oxford.

Visala, S. (1991). Broadening the empirical framework of information systems research. In Nissen,H., Klein, H. K., and Hirschheim, R. A. (eds.), Information Systems Research: ContemporaryApproaches and Emergent Traditions, North-Holland, Amsterdam.

Wade, S. (1991). A new course in systems analysis and design. Int. J. Inform. Manage. 11(3),238–247.

Walsham, G. (1993). Interpreting Information Systems in Organizations, John Wiley & Sons, Chich-ester.

Watson, H., Wood-Harper, T., and Wood, B. (1995). Interpreting methodology under erasure:Between theory and practice. Syst. Pract. 8(4), 441–468.

Watson, R., and Smith, R. (1988). Applications of the Lancaster soft systems methodology in Aus-tralia. J. Appl. Syst. Anal. 15, 3–26.

Willmott, H. (1989). OR as a problem situation: From soft systems methodology to critical systems.In Jackson, M. C., Keys, P., and Cropper, S. A. (eds.), Operations Research and the SocialSciences, Plenum, New York, pp. 65–78.

Wilson, B. (1990). Systems: Concepts, Methodologies and Applications, 2nd ed., John Wiley & Sons,Chichester.

Wood, J. R. G. (1992). Linking soft systems methodology and information systems. Systemist 14(3),133–135.

Wood, J. R. G., and Doyle, K. G. (1989). Doing the right thing right: An exploration between softsystems methodology and Jackson System Development. Proceedings of 33rd Annual Meetingof ISGR Conference, Program on Information Systems, Edinburgh, 2–7 July.

Wood-Harper, A. T., and Fitzgerald, G. (1982). A taxonomy of current approaches to systems anal-ysis. Comput. J. 25(1), 12–16.

Yolles, M. I. (1994). Generic metamodelling. Systemist 16(4), 269–286.

open research onlineoro.open.ac.uk/172/1/spar_holwell.pdf · soft systems methodology: other voices...

Documents